Research Talk

Reducing the number of variables in optimisation problems is one of key

approaches to reducing the problem to a manageable size, and often brings up

important extra insights into the problem at hand. There are settings where

one can remove exponentially many variables. This is instrumental for

packing problems, e.g. such that arise in error-correcting codes and other

similar settings such as crossing numbers of networks and graphs.

We review a number of our results in this area, both theoretical and

implemented in software, some of them relying on group representation

theory. Time allowing, we will touch upon questions of

reliability and reproducibility of computer-aided theoretical results,

and numeric robustness of linear optimisation.

Biography

Dmitrii (Dima) Pasechnik obtained his PhD in pure mathematics in 1995 from the

University of Western Australia, and has held postdoctoral and

research jobs in several universities in the Netherlands and Germany. In

2006-2013 he was an Assistant Professor at the newly established Division of

Mathematical Sciences of Nanyang Technological University (Singapore).

Then he moved to Oxford, where he held research positions at the CS

Department, and tutored mathematics at Pembroke College, Oxford University. At

present he is a lecturer at the CS Department, Oxford University .

Dmitrii's research interests are in combinatorics, optimization, algebra,

and in theory and practice of symbolic computing--he is involved in

development of SageMath, GAP, Macaulay2 computer algebra systems, and

related software. An adjacent area of his interests is

reproducibility of mathematical computations and constructions.

He supervised 3 PhD and 6 MSc students, published over 80 papers in

international journals and conference proceedings, and held a number of

grants in EU and Singapore.